Download Neonatal Ethanol Exposure Impairs Trace Fear Conditioning and

ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH
Vol. 40, No. 2
February 2016
Neonatal Ethanol Exposure Impairs Trace Fear
Conditioning and Alters NMDA Receptor Subunit
Expression in Adult Male and Female Rats
Molly J. Goodfellow, Khalid A. Abdulla, and Derick H. Lindquist
Background: Ethanol (EtOH) exposure in neonate rats during a period comparable to the human
third trimester, postnatal days (PD) 4 to 9, leads to persistent deficits in forebrain-dependent cognitive
function—modeling the dysfunction seen in individuals diagnosed with fetal alcohol spectrum disorders. EtOH-exposed adult rats are impaired in auditory trace fear conditioning (TFC), a form of Pavlovian conditioning in which a neutral conditioned stimulus (CS; tone) is followed by an aversive
unconditioned stimulus (US; footshock), with both stimuli separated in time by a stimulus-free “trace”
interval (TI). TFC acquisition depends on N-methyl-D-aspartate NMDA receptor (NMDAR) activation in the dorsal hippocampus (DH), ventral hippocampus (VH), and medial prefrontal cortex
(mPFC).
Methods: Male and female rat pups were sham-intubated (SI) or intragastrically intubated with
EtOH (5E; 5 g/kg/d) over PD 4 to 9 and, as adults, submitted to TFC with a 15-second tone CS and
30-second TI. Whole-cell tissue lysates from the DH, VH, and mPFC of TFC rats and DH synaptic/extrasynaptic membrane fractions from experimentally na€ıve animals were analyzed via Western blot for
NMDAR subunit (GluN1, GluN2A, GluN2B) expression.
Results: Freezing behavior during CS-alone test trials was significantly reduced in both male and
female 5E rats, relative to same-sex controls. Western blot results based on DH tissue samples revealed
a greater proportion of GluN2A to GluN2B subunits in 5E rats, relative to SI rats, and significantly
reduced synaptic GluN2B and PSD-95 expression.
Conclusions: EtOH-induced changes in DH NMDAR subunit expression—particularly synaptic
GluN2B, which is critical for TFC—are proposed to weaken long-term memory consolidation and,
during behavioral testing, diminish CS-evoked freezing behavior.
Key Words: Fetal Alcohol Spectrum Disorders, Trace Fear Conditioning, NMDA Receptors,
Hippocampus, Memory Consolidation.
F
ETAL ALCOHOL SPECTRUM disorders (FASD)
encompass a range of persistent physical and cognitive
abnormalities that can occur following gestational alcohol
exposure (Streissguth, 2007). Neuroimaging studies have
revealed structural and functional abnormalities in the hippocampus and prefrontal cortex of children with FASD
(Davis et al., 2011), contributing to the impaired performance seen in both children and adults with FASD when
tested in a variety of forebrain-dependent cognitive (e.g.,
learning and memory) tasks (Mattson et al., 2011). Ethanol
(EtOH) exposure over postnatal days (PD) 4 to 9 in rodents,
From the Department of Psychology (MJG, KAA, DHL), The Ohio
State University, Columbus, Ohio; and Department of Neuroscience
(KAA, DHL), The Ohio State University, Columbus, Ohio.
Received for publication March 14, 2015; accepted November 4, 2015.
Reprint requests: Derick H. Lindquist, PhD, Departments of
Psychology & Neuroscience, The Ohio State University, Psychology
Building, Room 45, 1835 Neil Avenue, Columbus, OH 43210; Tel.:
614-292-2236; Fax: 614-688-4733; E-mail: [email protected]
Copyright © 2016 by the Research Society on Alcoholism.
DOI: 10.1111/acer.12958
Alcohol Clin Exp Res, Vol 40, No 2, 2016: pp 309–318
a period comparable to the human third trimester (Bayer
et al., 1993), also impairs forebrain neurodevelopment, producing significant reductions in dendritic spine density, neurogenesis, and long-term potentiation (LTP) (Hamilton
et al., 2010; Klintsova et al., 2007; Puglia and Valenzuela,
2010).
We recently reported significant learning and memory deficits in PD 4 to 9 EtOH-exposed (5E) rats, when tested as
adults (DuPont et al., 2014). Relative to controls, adult 5E
rats were impaired in auditory trace fear conditioning
(TFC), in which a 15-second tone conditioned stimulus (CS)
was followed, 30 seconds later, by a footshock unconditioned
stimulus (US). TFC depends on a spatially distributed neural
circuit that includes the medial prefrontal cortex (mPFC)
and ventral hippocampus (VH) (Czerniawski et al., 2012;
Gilmartin and Helmstetter, 2010). When the “trace” interval
(TI; time separating CS offset from US onset) is longer than
~10-seconds, the dorsal hippocampus (DH) is required as
well (Chowdhury et al., 2005).
The toxic effects of ETOH on the developing hippocampus rely, at least in part, on subject sex. Following PD 4
to 9 exposure, for example, Schreiber and Hunt (2013)
309
GOODFELLOW ET AL.
310
found that adolescents of both sexes and adult females,
but not adult males, were impaired in visual TFC with a
10-second TI. Conversely, Johnson and Goodlett (2002)
reported deficits in the Morris water maze, another DHdependent task, in adult males (but not females) exposed
to EtOH across PD 7 to 9. Such variability following postnatal EtOH exposure suggests the underlying mechanisms
likely vary by age of testing and the behavioral task
employed. In the current study, adult (~PD 70) control
and 5E rats were submitted to DH-dependent TFC (30second TI)—training in female rats was initiated during
proestrus, when estradiol and progesterone are at their
highest levels (Woolley and Schwartzkroin, 1998).
Forebrain N-methyl-D-aspartate receptors (NMDARs)
are heterodimers composed of 2 GluN1 and 2 GluN2 subunits, GluN2A or GluN2B. The GluN2 subunits primarily
determine the voltage dependence of the Mg2+ channel
block, receptor kinetics, and downstream signaling (Erreger
et al., 2005). Early in life, GluN2B-containing NMDARs,
which gate more calcium but open less often than GluN2ANMDARs (Erreger et al., 2005; Monyer et al., 1994), are
ubiquitously expressed. Across postnatal development,
GluN2A expression is up-regulated and GluN2B expression
is slightly down-regulated (Monyer et al., 1994), producing a
more “stable” phenotype that requires greater excitation to
induce synaptic plasticity (Yashiro and Philpot, 2008). Aberrant NMDAR subunit expression has been documented in
preweanling rats exposed to EtOH during the pre- and/or
postnatal period (Nixon et al., 2002, 2004), as well as in adult
mice following prenatal exposure (Samudio-Ruiz et al.,
2010).
In the DH and mPFC, NMDARs play an important role
in the encoding and consolidation of TFC. For instance, pretraining infusions of an NMDAR antagonist into either
region impairs TFC, as evidenced by reduced freezing behavior during context and CS retention testing (Gilmartin and
Helmstetter, 2010; Quinn et al., 2005). DuPont and colleagues (2014) also found significant reductions in phosphorylated extracellular signal-related kinase (ERK)1/2
expression in CA1 and CA3 neurons from 5E rats sacrificed
1 hour after TFC, which was interpreted to reflect abnormal
NMDAR-dependent plasticity. It has not been previously
established, however, whether postnatal EtOH exposure
induces long-lasting changes in forebrain NMDAR subunit
expression when measured in adult rats. In addition to
NMDAR subunit composition, the location of membranebound GluN2A- and GluN2B-containing NMDARs (i.e.,
synaptic vs. extrasynaptic compartments) can also influence
learning-dependent synaptic plasticity (Hardingham and
Bading, 2010). Altogether, this study was designed to evaluate DH-dependent TFC in adult control and 5E rats (of both
sexes) and quantify the expression of GluN1, GluN2A, and
GluN2B subunits in whole-cell lysates (WCLs) from the DH,
VH, and mPFC and, separately, in DH synaptic membrane
fractions (SMFs) and extrasynaptic membrane fractions
(EMFs).
MATERIALS AND METHODS
Subjects and Neonatal Treatment
Long-Evans rats were bred and housed in The Ohio State
University vivarium and submitted to EtOH treatment as previously described (Goodfellow and Lindquist, 2014). On PD 3, litters were culled to 10 to 12 pups and paw-marked for
identification. Across PD 4 to 9, pups were administered EtOH
(11.33%) in milk solution via intragastric intubation twice daily,
2 hours apart (5 g/kg/d; 5E), followed 2 hours later by milk alone.
Sham-intubated (SI) control rats received identical treatment
except no milk or EtOH was given. On PD 4, blood samples for
determining blood alcohol concentration (BAC) were taken via tail
clip just prior to the milk-alone intubation. Samples from the SI
rats were discarded; the 5E samples were centrifuged and plasma
separated. Peak BACs were determined using an Analox GL5
Analyzer (Analox Instruments, Lunenburg, MA), which measures
the rate of oxygen consumption resulting from oxidation of EtOH
in the sample.
Rats were weaned on PD 21, same-sex group housed until PD
60 and, consistent with our previous studies, individually housed
until sacrificed (DuPont et al., 2014; Goodfellow and Lindquist,
2014). Testing occurred between PD 65 to 75 around midday
(~1100 to 1500) during the light phase (12-hour light/dark cycle—
lights on at 0600). Prior work has demonstrated that SI and
unhandled/unintubated control rats exhibit similar patterns of
freezing behavior during TFC and retention testing with TIs ranging from 5 to 30-seconds (DuPont et al., 2014; Hunt et al., 2009).
Considering SI rats are the critical comparison group for 5E rats
(Schreiber and Hunt, 2013), and to conserve animals, only SI and
5E rats were used in the current study. A total of 89 rats were
used; in cases where more than 1 male or female per treatment
group per litter were used (SI n = 8 males, 6 females; 5E n = 4
males, 8 females), data were averaged (i.e., 26 rats yielded 13 data
points), resulting in N = 76.
Estrous Cycle Tracking
From PD 61 to 65, at ~0930 females were vaginally swabbed
with a sterile cotton swab soaked in 0.9% saline. Samples were
applied to a subbed slide and stained with toluidine blue to determine stage of estrous based on cell morphology, as described in
Shors (1998). Those in proestrus underwent TFC 2 to 4 hours
later or, separately, were sacrificed to quantify membrane-bound
NMDAR protein expression. Females not in proestrus continued
to be swabbed daily until they entered proestrus. As a control,
male genitals were also swabbed until 1 male could be conditioned
(or sacrificed) at the same time as 1 female littermate. For all
females, stage of estrous was numerically coded (1 = estrus,
2 = diestrus 1, 3 = diestrus 2, and 4 = proestrus) and averaged
from PD 61 to 64. A normal estrous cycle would result in a mean
of 2.5.
Apparatus
Standard operant boxes (Coulbourn Instruments, Allentown,
PA) in sound-attenuating chambers were composed of 2 stainless
steel and 2 plexiglass walls, and a floor with 0.5-cm stainless steel
bars spaced 1.5 cm apart. A shock generator (model 82400; Lafayette Instruments, Lafayette, IN) and neon grid scrambler (model
58020; Lafayette Instruments) delivered the footshock US. The interior was illuminated with infrared lights, and freezing behavior was
recorded with a video camera (Model WDSR-2005SC; Circuit Specialists, Inc., Mesa, AZ) mounted to the top of the conditioning
chamber. The video signal was inputted to FreezeScan (CleverSys,
Inc., Reston, VA), which quantified freezing behavior within experimenter-defined time bins.
NEONATAL ETHANOL ALTERS NMDA RECEPTORS
311
Trace Fear Conditioning
Rats were carried in their home cages, 2 at a time, to a well-lit
testing room and placed in an operant box illuminated by a 15W bulb and scented with a vinegar/water (1:5) solution. After a
240 30-second baseline, rats were presented with a 15-second,
2.8 kHz, 80 dB tone CS followed 30 seconds later by a 1-second,
0.8 mA footshock US, resulting in a 45-second interstimulus
interval (ISI; time from CS onset to US onset). In all, rats
received 10 CS–US presentations, separated by an intertrial interval (ITI) of 240 30-seconds. Approximately 24 and 48 hours
later, rats underwent context and CS-alone testing in counterbalanced order.
Context fear was assessed in the conditioning chamber. After a
120-second baseline, freezing was measured for 600-seconds and
the rat removed 120-seconds later. CS-evoked freezing was
assessed in a novel context. Rats were wheeled into a dark testing
room (lit only with a red overhead light) on a metal cart, with
each cage covered by a towel. The chamber was scented with WindexÒ (S.C. Johnson & Son, Inc., Racine, WI), a gray opaque plexiglass sheet covered the grid bars, an interior fan provided 60 dB
white noise, a small magnet was placed on 1 metal wall, and a
pink geometric figure was attached to the door. Freezing was measured for 1 minute prior to the first tone CS to verify the context
switch was successful. The 15-second tone CS was presented only
5 times (90 10-second ITI) to minimize extinction learning.
Freezing was recorded during each CS-alone trial and the 30-second period following CS offset (i.e., the TI). Rats were returned to
the vivarium 120-seconds after the final trial.
Whole-Cell Lysate Tissue Preparation
At least 3 days after behavioral testing, rats were deeply anesthetized with isoflurane and decapitated. Brains were removed and
the DH, VH, and mPFC rapidly dissected and stored at 80°C.
While TFC requires the prelimbic, but not infralimbic, region of the
mPFC (Gilmartin and Helmstetter, 2010), we were unable to confidently separate them—thus, prelimbic and infralimbic cortices were
analyzed together. Tissue was homogenized in 95°C buffer (1%
sodium dodecyl sulfate [SDS], 10 mM Tris [pH 7.4], 1 mM
Na3VO4) and centrifuged at 29,0009g at 15°C for 10 minutes. The
supernatant (WCL) was stored at 80°C.
Membrane Fraction Tissue Preparation
Potential sex differences in NMDAR subunit expression were
examined in experimentally na€ıve rats under the same hormonal
conditions (i.e., proestrus for female rats) as at the start of TFC.
SMFs and EMFs were isolated via a protocol adapted from
Samudio-Ruiz and colleagues (2010) and Goebel-Goody and colleagues (2009). Tissue was lysed in homogenization buffer
(20 mM Tris, pH 7.4, 1 mM EDTA, 320 mM sucrose, 20 mM
Na4P2O7, 10 mM NaF, 20 mM b-glycerophosphate, 0.2 mM
Na3VO4) and centrifuged (1,0009g, 10 minutes). The supernatant
was centrifuged again (15,0009g, 30 minutes), and the resulting
pellet was homogenized with distilled/deionized cold water containing protease inhibitor cocktail (Sigma-Aldrich, St. Louis,
MO). HEPES-NaOH buffer (pH 7.4, 7.5 mM) was added and
kept on ice 30 minutes before centrifugation (22,0009g, 20 minutes). Pellets were resuspended in Triton X-100 buffer (1% Triton, 10 mM Tris, pH 7.4, 5 mM NaF, 1 mM EDTA, 0.5 mM
EGTA, 0.2 mM Na3VO4), kept on ice for 30 minutes and centrifuged (100,0009g, 60 minutes) to yield the triton-insoluble pellet (SMF). Acetone was added to the supernatant, incubated
4 hours ( 20°C), and centrifuged (15,0009g, 10 minutes) to
obtain the Triton-soluble (EMF) pellet. The EMF and SMF were
each combined with homogenization buffer containing 1% SDS,
heated (95°C, 5 minutes), and stored at 80°C. Bradford assays
determined protein concentrations (ThermoScientific, Rockford,
IL); due to low yields, total protein loaded and antibody concentrations were titrated prior to data collection (see Table 1).
Western Blotting
Samples containing 20 lg (WCL), 7 lg (SMF), or 5 lg (EMF)
protein were combined 1:1 with loading buffer and heated (95°C,
5 minutes). Samples were loaded into a 7.5% TGX gel (Bio-Rad,
Hercules, CA), electrophoresed (150V, ~70 minutes), and transferred to a nitrocellulose membrane (Bio-Rad) at 100V for 60 minutes. Membranes were cut to separate GluN2A or GluN2B from
GluN1, PSD-95, and the loading control (GAPDH or actin) and
equilibrated in 5% nonfat dry milk (NFDM; Bio-Rad) for 60 minutes at room temperature. Membranes incubated overnight at 4°C
in primary antibody in NFDM. The next day, membranes were
rinsed in TBST 4 9 7 minutes and equilibrated in secondary antibodies in NFDM for 60 minutes. Membranes were rinsed in TBST
4 9 7 minutes and covered in Clarity Western Blot Substrate (BioRad) for 5 minutes. Blots were imaged with a FluorChem M system
and analyzed with AlphaView software (ProteinSimple, Santa
Clara, CA).
Data Analysis
Data were analyzed via 1-way, multifactorial, and repeated-measures analysis of variance (ANOVA), and (when required) Bonferroni-corrected 1-way ANOVAs. Tukey’s post hoc tests were used to
study significant main effects and interactions. A significant post
hoc test implies p < 0.05. As determined prior to the start of the
experiment, behavioral data were excluded if it was >2 standard
deviations from the group mean (see Broadwater and Spear, 2013;
Dokovna et al., 2013)—2 5E females, 2 5E males, and 1 SI female
were dropped.
Table 1. Antibodies Used for Western Blotting Including Their Host Animal, Molecular Weight, Manufacturer, and the Concentrations Used for WholeCell Lysate (WCL), Synaptic Membrane Fraction (SMF), and Extrasynaptic Membrane Fraction (EMF)
Target
Host
Molecular weight
Manufacturer
WCL concentration
SMF concentration
EMF concentration
GluN2A
GluN2B
GluN1
PSD-95
Actin
GAPDH
Rabbit
Mouse
Rabbit
Rabbit
Mouse
Rabbit
Mouse
Mouse
Goat
Goat
180 kD
180 kD
120 kD
95 kD
42 kD
38 kD
–
–
PhosphoSolutions, Aurora, CO
Millipore, Billerica, MA
BD Biosciences, San Jose, CA
Cell Signaling, Danvers, MA
PhosphoSolutions, Aurora, CO
Novus, Littleton, CO
Pierce, Rockford, IL
BD Biosciences, San Jose, CA
1:3,000
1:1,500
1:3,000
–
–
1:5,000
1:5,000
1:2,000
1:1,000
1:2,000
1:1,250
1:1,000
1:1,500
–
1:5,000
1:2,000
1:200
1:400
1:250
1:1,000
1:1,500
–
1:5,000
1:2,000
GOODFELLOW ET AL.
312
RESULTS
Blood Alcohol Concentration
Peak BACs (mean SE) in 5E rats were 351.7 6.2 mg/
dl in females and 344.5 7.2 mg/dl in males—no significant
sex differences were found (p = 0.45).
Estrous Cycle
Based on our numerical coding, normal cycling would
yield a value of 2.5. Treatment group means (SE) were
2.36 0.09 in SI rats and 2.24 0.07 in 5E rats. A 1-way
(Treatment) ANOVA failed to find significant treatment
group differences (p = 0.31; data not shown), suggesting the
SI and 5E female rats cycled similarly prior to TFC. While
the estrous cycle was not tracked following TFC, the normal
cycling seen between treatment groups suggests all females
were likely in estrus or diestrus 1 during the 2 test sessions.
Trace Fear Conditioning
Previous research from our laboratory demonstrated no
differences in CS-evoked freezing behavior between control
and 5E treatment groups during the TFC session (DuPont
et al., 2014), indicating intact sensory/motor responding and
associative learning for both groups. The current study
expands on our previous results by comparing SI and 5E
freezing percentages during the 15-second tone CS and,
following its offset, the 30-second TI. Context-dependent
freezing was measured during a separate test. The 2
(Treatment) 9 2 (Sex) ANOVAs revealed significant Treatment group main effects for the CS, F(1, 31) = 15.19,
p < 0.001, TI, F(1, 31) = 8.90, p < 0.01, and context, F(1,
31) = 9.75, p < 0.01. Post hoc analyses indicate the 5E rats
froze less than SI rats during all 3 test phases. A significant
Sex effect was found for context only, F(1, 31) = 5.13,
p < 0.05, with females freezing less than males, regardless of
neonatal treatment. The Treatment 9 Sex interaction was
not significant for any test phase.
Mean freezing rates in males and females were separately
analyzed across all 3 test phases with 1-way (Treatment)
ANOVAs. In females (Fig. 1A), there was a significant main
effect for the tone CS, F(1, 15) = 10.47, p < 0.01, and TI, F
(1, 15) = 7.24, p < 0.05, with 5E females impaired relative to
SI females. Significant treatment group differences for context test freezing just missed significance (p = 0.06). In males
(Fig. 1B), there were significant main effects for the tone
CS, F(1, 16) = 5.71, p < 0.05, and context, F(1, 16) = 6.07,
p < 0.05, but not to the TI (p = 0.12). The 5E rats were again
impaired relative to SI rats.
Due to the relatively modest sex differences, behavioral
data were subsequently collapsed across sex and analyzed via
1-way (Treatment) ANOVAs. Results revealed significant
main effects for the tone CS, F(1, 33) = 15.48, p < 0.001, TI,
F(1, 33) = 8.32, p < 0.01, and context, F(1, 33) = 8.55,
p < 0.01, with 5E rats freezing less than SI rats (Fig. 2A).
The significant context freezing impairment differs from our
previous study (DuPont et al., 2014) and is likely attributable to the larger group sizes and higher power in the current
study. Next, 2 (Treatment) 9 5 (Trial) repeated-measures
ANOVAs were used to examine freezing during and following each CS-alone trial or, for the context test, across five 2minute bins. As seen in Fig. 2B, the low level of freezing in
both treatment groups during the pre-tone period in the
novel context indicates a successful context switch (i.e., little
context fear generalization). Across the 5 CS presentations,
results revealed significant effects for Trial, F(4, 132) = 2.49,
p < 0.05, and the Treatment 9 Trial interaction, F(4,
132) = 3.06, p < 0.05. Separate 1-way ANOVAs on each
trial found that 5E rats froze significantly less during trials 3
to 5. During the TI (Fig. 2C), a significant Treatment 9 Trial interaction was also found, F(4, 132) = 3.19,
Fig. 1. Percent freezing at test (mean SE), by treatment group and sex. (A) Female ethanol-exposed (5E) rats were impaired relative to female
sham-intubated (SI) rats in freezing to the tone conditioned stimulus (CS) and trace interval (TI). (B) The 5E male rats were also impaired relative to SI
male rats in freezing to the tone CS and training context. An asterisk (*) denotes a significant treatment group effect (p < 0.05).
NEONATAL ETHANOL ALTERS NMDA RECEPTORS
313
Fig. 2. Percent freezing at test (mean SE), collapsed across sex. (A) Freezing was significantly reduced in ethanol-exposed (5E) rats, relative to
sham-intubated (SI) rats, to the tone conditioned stimulus (CS), trace interval (TI), and training context. (B) No treatment group differences were seen in
freezing prior to the first CS-alone trial (pre-tone period), indicating a successful context switch. Across the 5 CS-alone trials, the 5E rats froze significantly
less over trials 3 to 5. (C) Following CS offset, freezing was measured during the 30-second TI. The 5E rats froze significantly less on trials 3 and 4. (D)
Across the 10-minute context test (shown in 2-minute bins), the 5E rats were significantly impaired relative to SI rats. An asterisk (*) denotes a significant
treatment group effect; a dollar sign ($) represents a significant trial effect; an exclamation point (!) indicates a significant Treatment 9 Trial interaction
(p < 0.05).
p < 0.05, with 5E rats significantly impaired during trials 3
and 4. For the context test (Fig. 2D), only the Trial main
effect was significant, F(4, 132) = 7.48, p < 0.001, with both
groups freezing less over time.
Finally, we sought to characterize freezing behavior across
the entire 45-second ISI (15-second CS + 30-second TI;
Fig. 3). Freezing was contrasted between treatment groups
during each of nine 5-second bins (CS: bins 1 to 3, TI: bins 4
to 9). A repeated-measures ANOVA revealed significant
main effects for Treatment, F(1, 33) = 10.83, p < 0.01, and
Time, F(8, 264) = 9.31, p < 0.001, but not their interaction
(p = 0.77). Post hoc testing verified the 5E rats were impaired
relative to SI rats. Bonferroni-corrected 1-way ANOVAs
were applied to each of the nine 5-second bins—requiring (at
9 contrasts) p < 0.0056 for statistical significance, maintaining a family-wise a = 0.05. Results indicate that 5E rats froze
significantly less than SI rats across bins 2 and 3 (the final 10seconds of the tone CS) and bin 4 (first 5-seconds of the TI).
As seen in Fig. 3, the 5E rats were clearly impaired relative
to SI rats but the pattern of freezing was similar for both
groups across the averaged 45-second trial.
Whole-Cell Lysate Immunoblots
Following the completion of testing, brain tissue was analyzed for NMDAR subunit expression via Western blotting.
Note, tissue was lost due to a 80°C freezer malfunction,
reducing the sample size of both treatment groups. The optical density (OD) values of individual subunits were first normalized to OD values of loading control expression ([target
OD/loading control OD] 9 loading control mean). As we
did not control for estrous cycle on the day of sacrifice, data
were collapsed across sex and analyzed via 1-way (Treatment) ANOVAs in the DH, VH, and mPFC (SI n = 3 males,
4 females; 5E n = 3 males, 3 females). Importantly,
immunoreactivity for the GAPDH loading control did not
GOODFELLOW ET AL.
314
increase in the proportion of GluN2A expression relative to
GluN2B in 5E rats, F(1, 11) = 4.86, p = 0.05 (Fig. 4D). The
difference cannot be specifically attributed to TFC, however.
All animals were sacrificed at least 3 days following the completion of testing.
Subcellular Fraction Immunoblots
Fig. 3. Averaged percent freezing (mean SE) over the 45-second
interstimulus interval (ISI; 15-second conditioned stimulus + 30-second
trace interval). Each data point (5-second bin) is the average of 5 test trials.
Ethanol-exposed (5E) rats froze significantly less than sham-intubated (SI)
rats across the whole ISI. Post hoc testing indicates the 5E rats froze significantly less than SI rats across bins 2 to 4 (seconds 6 to 20). An asterisk
(*) denotes a significant treatment group effect and a dollar sign ($) indicates a significant effect of time (p < 0.05).
differ by treatment group. No significant treatment group
differences were found in GluN2A, GluN2B, or GluN1
expression in any brain region (Fig. 4A–C). Closer examination of the DH results, however, revealed a significant
For subcellular fractions, experimentally na€ıve females
were sacrificed during proestrus along with a male littermate.
Expression of the loading control, actin, did not significantly
differ between treatment groups or sex in any region or fraction. Treatment 9 Sex ANOVAs applied to DH synaptic
fractions (SI n = 6 males, 6 females; 5E n = 5 males, 6
females) and DH extrasynaptic fractions (SI n = 5 male, 5
females; 5E n = 4 males, 4 females) failed to uncover significant effects for any protein. For the main effect of Sex, all pvalues were >0.40. The data suggest NMDAR subunits are
expressed at comparable levels in male and female adult rats.
Immunoblot data were thus collapsed across sex and analyzed via 1-way (Treatment) ANOVAs. For DH synaptic
fractions, significant treatment group differences were
seen for GluN2B, F(1, 21) = 4.42, p < 0.05, and PSD-95,
F(1, 21) = 4.46, p < 0.05, with expression decreased in 5E
rats relative to SI rats (Fig. 5A). The Treatment main effect
failed to reach statistical significance for GluN1 (p = 0.10) or
GluN2A (p = 0.20). VH synaptic fractions were also analyzed, with no significant Treatment group differences seen
Fig. 4. Western blot optical density (OD; mean SE), normalized to GAPDH OD, for N-methyl-D-aspartate receptor (NMDAR) subunits. Results are
based on whole-cell lysates from ethanol-exposed (5E) and sham-intubated (SI) rats taken from the (A) dorsal hippocampus (DH), (B) ventral hippocampus (VH), and (C) medial prefrontal cortex (mPFC). No treatment group differences were noted for NMDAR subunit expression in any brain region. (D)
Ratio of GluN2A relative to GluN2B by region—a significant increase in the GluN2A/GluN2B ratio was found in DH only. (E) Representative DH immunoblot blot images. An asterisk (*) denotes a significant treatment group effect (p < 0.05).
NEONATAL ETHANOL ALTERS NMDA RECEPTORS
315
Fig. 5. Western blot optical density (OD; mean SE), normalized to actin OD, for N-methyl-D-aspartate receptor(NMDAR) subunits. Results are
based on (A) dorsal hippocampus (DH) synaptic membrane fractions and (B) DH extrasynaptic membrane fractions, taken from ethanol-exposed (5E)
and sham-intubated (SI) rats. In 5E rats, GluN2B and PSD-95 were significantly reduced in the synaptic membrane fraction only. (C) The proportion of
GluN2A relative to GluN2B did not significantly differ between treatment groups in either membrane fraction. (D) Representative synaptic and extrasynaptic GluN2B immunoblots—note the absence of PSD-95 expression in the extrasynaptic fractions. An asterisk (*) denotes a significant treatment group
effect (p < 0.05).
for any protein (data not shown). Figure 5B shows results
from DH extrasynaptic fractions, which are characterized by
a lack of PSD-95 expression (Fig. 5D). One-way (Treatment)
ANOVAs did not uncover significant effects for any protein
of interest. Interestingly, analysis of the GluN2A/GluN2B
ratio found no treatment group differences for DH synaptic
fractions (p = 0.69), while a modest trend toward statistical
significance (p = 0.11) was seen with the extrasynaptic fractions (Fig. 5C). Protein levels from extrasynaptic VH and
both synaptic and extrasynaptic mPFC were insufficient for
analysis.
detected biochemically. A significant increase in the
GluN2A/GluN2B ratio (based on WCL immunoblots) was
seen in the DH, but not the VH or mPFC, of 5E rats (Fig. 4).
Treatment group differences were also seen in DH synaptic
but not extrasynaptic fractions, with synaptic GluN2B and
PSD-95 significantly reduced in 5E rats (Fig. 5). Changes in
NMDAR subunit composition and receptor kinetics are proposed to impede the maintenance of TFC-dependent synaptic plasticity, restricting the consolidation of long-term trace
fear memories.
Trace Fear Conditioning
DISCUSSION
In the current study, TFC performance and Western blot
analyses of NMDAR subunit expression in the DH, VH,
and mPFC were used to study the deleterious effects of postnatal binge-like EtOH exposure in adult subjects. Following
TFC, the 5E rats demonstrated significant freezing deficits to
the CS, TI, and context (Figs 2 and 3). Only modest sex differences were seen in behavior (Fig. 1), and none were
Forebrain cognitive function was assessed in adult rats via
TFC, a Pavlovian paradigm that depends on higher order
cognition, including attention-dependent associative learning
and working memory (Raybuck and Lattal, 2014). During
retention testing, the 5E rats froze significantly less than SI
rats during the tone CS and ensuing TI in a novel context,
and when re-exposed to the conditioning context (Fig. 2).
The results are consistent with prior research documenting
GOODFELLOW ET AL.
316
impaired forebrain-dependent associative learning following
PD 4 to 9 EtOH administration (DuPont et al., 2014; Schreiber and Hunt, 2013) and support the proposition that earlylife EtOH exposure can induce persistent deficits in cognitive
function. The effect is likely not due to treatment group differences in somatic development—adult control and 5E rats
do not differ in terms of body weight (DuPont et al., 2014;
Goodfellow and Lindquist, 2014).
Results indicate that, following TFC, 5E rats of both
sexes froze significantly less during CS-alone test trials
(Figs 1 and 2). In comparison, Schreiber and Hunt (2013)
reported that adult females, but not males, exposed to
EtOH over PD 4 to 9 were significantly impaired in CSevoked freezing following visual TFC. There are a number
of methodological differences between the 2 studies—for
example, Schreiber and Hunt (2013) used freely cycling
females, a light CS, a 10-second TI, and group-housed animals throughout testing—that likely contribute to the disparate results. In the current study, 5E female rats also
froze less during the TI (but not in the training context) relative to SI females, whereas 5E male rats froze less in the
training context (but not during the TI) relative to SI males
(Fig. 1). That said, the sex effects were still modest—both
male and female 5E rats demonstrated a clear trend toward
reduced freezing during the TI and in the training context
relative to same-sex controls.
DuPont and colleagues (2014) found significant reductions in 5E rats’ freezing behavior to the tone CS following
TFC with a 15- and 30-second TI, but not a 5-second TI.
As previously discussed, the mPFC seems to be required
for TFC across all TIs, whereas the DH is only engaged
when the TI exceeds ~10-seconds (Chowdhury et al., 2005).
Such results suggest the current TFC deficit seen in 5E rats
depends, principally, on DH dysfunction. To better relate
putative deficits in DH function to freezing behavior, we
broke down the 45-second ISI (15-second CS + 30-second
TI) into nine 5-second bins. When freezing was averaged
across all 5 CS-alone test trials, both SI and 5E rats showed
increased freezing as the time of the expected US
approached (Fig. 3). Despite the 5E rats freezing less overall, both treatment groups demonstrated a similar pattern
of freezing behavior across the 45-second ISI. Results support our hypothesis that the memory retrieved by 5E rats
at the time of testing is weaker (i.e., less consolidated) than
the memory retrieved by SI rats (Goodfellow and Lindquist, 2014).
TFC acquisition relies on GluN2B-containing NMDARs
in both the mPFC (prelimbic cortex) and DH. Pretraining
administration of a GluN2B-specific antagonist (Ro25-6981)
into either region selectively blocks TFC acquisition, while
delay fear conditioning (in which the CS and US overlap and
coterminate) and context fear conditioning are unaffected
(Gao et al., 2010; Gilmartin et al., 2013). DuPont and colleagues (2014) also reported significant reductions in DH
phosphorylated ERK1/2 in 5E rats, which depends on the
activation of GluN2B-containing NMDARs (Krapivinsky
et al., 2003). NMDAR-gated calcium flux stimulates the
translocation and binding of Ca2+/calmodulin-dependent
protein kinase II to the synaptic GluN2B c-terminus domain,
which leads to the activation of plasticity-related signaling
molecules essential to LTP maintenance (Lisman and
Raghavachari, 2015). Lower synaptic GluN2B expression in
5E rats is proposed to hinder LTP maintenance following
TFC, disrupting long-term memory consolidation of the
newly learned information.
NMDA Receptors
Western blot results based on WCLs revealed a significant
elevation in GluN2A subunits relative to GluN2B in the
DH, but not VH or mPFC, of adult 5E rats (Fig. 4D), which
has the potential to limit or perturb NMDAR-gated calcium
flux and synaptic plasticity in 5E rats (Shouval et al., 2002).
NMDAR subunit expression is known to be altered as a
function of development (Monyer et al., 1994), which may
explain prior studies that failed to find lasting differences in
NMDAR subunit expression following perinatal EtOH
exposure. In Nixon and colleagues (2002), for example, the
effects of PD 4 to 9 EtOH exposure on cortical and hippocampal NMDAR subunits were assessed on PD 10, 14,
and 21. Results from EtOH-exposed rats revealed a selective
increase in cortical GluN2A expression on PD 21 only. Zink
and colleagues (2011) argue that early-life EtOH exposure
may induce greater (perhaps cumulative) changes in
GluN2A and GluN2B gene expression when measured in
adults versus neonates. Current results suggest a similar
developmental trajectory for membrane-bound GluN2A and
GluN2B protein expression.
EtOH-induced changes in the estrous cycle could also contribute to observed changes in NMDAR subunit expression
and TFC performance. Proestrus is correlated with increases
in dendritic spine density and synapse number (Gould et al.,
1990; Woolley and McEwen, 1992), as well as a decrease in
the LTP induction threshold (Cordoba Montoya and Carrer,
1997). No differences were seen between SI and 5E female
rats in day-to-day cycling—however, we cannot rule out
potential treatment group differences in ovarian hormonal
levels. Estradiol has been previously shown, for instance, to
increase GluN1 translation in CA1 neurons of ovariectomized rats (Gazzaley et al., 1996) and to increase both
spine numbers and PSD-95 expression in the DH of ovariectomized mice (Li et al., 2004). Given that male and female
5E rats were both impaired in TFC and showed no differences in NMDAR subunit expression, we suggest the estrous
cycle and ovarian secretions are only minimally affected, if at
all, by postnatal EtOH exposure.
Results based on DH membrane fractions found significant treatment group differences in synaptic GluN2B and
PSD-95 expression only (Fig. 5A). Adult mice exposed to
EtOH before birth also had significant reductions in GluN2B
(though not PSD-95) in SMFs (Samudio-Ruiz et al., 2010).
In contrast, Kervern and colleagues (2015) reported an up-
NEONATAL ETHANOL ALTERS NMDA RECEPTORS
regulation in synaptic GluN2B protein expression following
combined pre- and postnatal EtOH exposure, when measured in young adult rats. The reason for the discrepant
results is unclear—the extended period of EtOH exposure in
Kervern and colleagues (2015), and their focus on area CA1,
likely contribute to the different levels of protein expression
seen between studies. While only GluN2B and PSD-95 were
significantly reduced in 5E rats, all synaptic NMDAR subunits showed a general trend toward reduced expression, relative to SI rats. Potentially contributing to these effects,
postnatal EtOH exposure is known to induce apoptosis,
albeit cell loss in adult animals is typically restricted to area
CA1, while areas CA3 and the dentate gyrus seem more
resistant to EtOH’s toxic effects (Bonthius and West, 1991;
Tran and Kelly, 2003). Moreover, perinatal EtOH exposure
has also been reported, in PD 70 rats, to significantly reduce
the density of synapses within the hippocampus (Kuge et al.,
1993). Notwithstanding the widespread consequences of
early-life EtOH exposure on hippocampal neuronal and
synaptic development, the significant (and selective) reduction in synaptic GluN2B is proposed to underlie, at least in
part, the TFC deficits observed in 5E rats.
NMDARs are localized within synaptic and extrasynaptic
compartments, which governs the receptors coupling to
specific intracellular signaling pathways (Hardingham,
2006). WCL immunoblot results revealed no significant
treatment group differences for any NMDAR subunit suggesting intracellular stores are not significantly altered by
postnatal EtOH administration, although NMDAR trafficking to and from the synapse could be disrupted (SamudioRuiz et al., 2010). Extrasynaptic NMDARs are principally
activated under conditions of high glutamate release, which
induces opposing effects to synaptic NMDAR activation
(Ivanov et al., 2006). As discussed above, synaptic
NMDARs play a well-documented role in LTP induction
and maintenance (Lisman and Raghavachari, 2015; Lu
et al., 2001), and postnatal EtOH is known to inhibit LTP
(Puglia and Valenzuela, 2010). Extrasynaptic NMDARs, on
the other hand, are thought to dephosphorylate ERK1/2
and promote long-term depression (Ivanov et al., 2006; Liu
et al., 2013). Thus, the TFC impairments seen in 5E rats
may be attributable to impaired LTP and/or an increase in
LTD mediated through extrasynaptic NMDARs, whose
expression was less affected by postnatal EtOH than synaptic NMDARs. It is not clear, however, to what degree
extrasynaptic NMDARs are activated in adult 5E rats—to
our knowledge, hippocampal glutamate release has not been
measured in older animals following perinatal EtOH exposure. Acutely, early-life EtOH exposure is reported to
rapidly suppress glutamatergic transmission (e.g., Mameli
et al., 2005). Additional research is required to better
characterize how learning-dependent synaptic plasticity is
altered in EtOH-exposed rodents, based on NMDAR
compartmental localization and the post translational modifications (e.g., phosphorylation) that occur downstream of
NMDAR activation.
317
CONCLUSIONS AND FUTURE DIRECTIONS
Current results extend our prior research and confirm the
disruptive effects of postnatal EtOH exposure on forebraindependent cognitive function in male and female adult subjects. Following TFC, the 5E rats froze significantly less
during the tone CS and following its offset, due, we suggest,
to EtOH-induced alterations in hippocampal NMDAR
subunit composition and function. The finding that synaptic GluN2B expression, in particular, is significantly
reduced in 5E rats is proposed to induce persistent deficits
in LTP maintenance and long-term memory consolidation.
Results are anticipated to further research aimed at
developing novel pharmacological treatments targeted at
glutamatergic signaling and plasticity, with the goal of
improving cognitive function in FASD rodent models
and, ultimately, children and adults exposed to gestational
alcohol.
ACKNOWLEDGMENTS
The authors thank Jennifer Coppola, Shelby Parsons, and
Gabriel Boyle for their assistance with running and analyzing
the behavioral experiments. We also thank Dr. Kevin Caldwell for his advice on subcellular fractionation and Dr.
Kevin Foust for use of his ultracentrifuge. Funding support
generously provided by ABMRF, The Foundation for Alcohol Research, grant 60034780. The authors declare no conflicts of interest.
REFERENCES
Bayer SA, Altman J, Russo RJ, Zhang X (1993) Timetables of neurogenesis
in the human brain based on experimentally determined patterns in the
rat. Neurotoxicology 14:83–144.
Bonthius DJ, West JR (1991) Permanent neuronal deficits in rats exposed to
alcohol during the brain growth spurt. Teratology 44:147–163.
Broadwater M, Spear LP (2013) Age differences in fear retention and extinction in male Sprague-Dawley rats: effects of ethanol challenge during conditioning. Behav Brain Res 252:377–387.
Chowdhury N, Quinn JJ, Fanselow MS (2005) Dorsal hippocampus involvement in trace fear conditioning with long, but not short, trace intervals in
mice. Behav Neurosci 119:1396–1402.
Cordoba Montoya DA, Carrer HF (1997) Estrogen facilitates induction of
long term potentiation in the hippocampus of awake rats. Brain Res
778:430–438.
Czerniawski J, Ree F, Chia C, Otto T (2012) Dorsal versus ventral hippocampal contributions to trace and contextual conditioning: differential
effects of regionally selective NMDA receptor antagonism on acquisition
and expression. Hippocampus 22:1528–1539.
Davis K, Desrocher M, Moore T (2011) Fetal alcohol spectrum disorder: a
review of neurodevelopmental findings and interventions. J Dev Phys Disabil 23:143–167.
Dokovna LB, Jablonski SA, Stanton ME (2013) Neonatal alcohol exposure
impairs contextual fear conditioning in juvenile rats by disrupting cholinergic function. Behav Brain Res 248:114–120.
DuPont CM, Coppola JJ, Kaercher RM, Lindquist DH (2014) Impaired
trace fear conditioning and diminished ERK1/2 phosphorylation in the
dorsal hippocampus of adult rats administered alcohol as neonates. Behav
Neurosci 128:187–198.
318
Erreger K, Dravid SM, Banke TG, Wyllie DJ, Traynelis SF (2005) Subunitspecific gating controls rat NR1/NR2A and NR1/NR2B NMDA channel
kinetics and synaptic signalling profiles. J Physiol 563:345–358.
Gao C, Gill MB, Tronson NC, Guedea AL, Guzman YF, Huh KH, Corcoran KA, Swanson GT, Radulovic J (2010) Hippocampal NMDA receptor
subunits differentially regulate fear memory formation and neuronal signal
propagation. Hippocampus 20:1072–1082.
Gazzaley AH, Weiland NG, McEwen BS, Morrison JH (1996) Differential
regulation of NMDAR1 mRNA and protein by estradiol in the rat hippocampus. J Neurosci 16:6830–6838.
Gilmartin MR, Helmstetter FJ (2010) Trace and contextual fear conditioning require neural activity and NMDA receptor-dependent transmission in
the medial prefrontal cortex. Learn Mem 17:289–296.
Gilmartin MR, Kwapis JL, Helmstetter FJ (2013) NR2A- and NR2B-containing NMDA receptors in the prelimbic medial prefrontal cortex differentially mediate trace, delay, and contextual fear conditioning. Learn
Mem 20:290–294.
Goebel-Goody SM, Davies KD, Alvestad Linger RM, Freund RK, Browning MD (2009) Phospho-regulation of synaptic and extrasynaptic Nmethyl-d-aspartate receptors in adult hippocampal slices. Neuroscience
158:1446–1459.
Goodfellow MJ, Lindquist DH (2014) Significant long-term, but not shortterm, hippocampal-dependent memory impairment in adult rats exposed
to alcohol in early postnatal life. Dev Psychobiol 56:1316–1326.
Gould E, Woolley CS, Frankfurt M, McEwen BS (1990) Gonadal steroids
regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J Neurosci 10:1286–1291.
Hamilton GF, Whitcher LT, Klintsova AY (2010) Postnatal binge-like alcohol
exposure decreases dendritic complexity while increasing the density of
mature spines in mPFC layer II/III pyramidal neurons. Synapse 64:127–135.
Hardingham GE (2006) 2B synaptic or extrasynaptic determines signalling
from the NMDA receptor. J Physiol 572:614–615.
Hardingham GE, Bading H (2010) Synaptic versus extrasynaptic NMDA
receptor signalling: implications for neurodegenerative disorders. Nat Rev
Neurosci 11:682–696.
Hunt PS, Jacobson SE, Torok EJ (2009) Deficits in trace fear conditioning in
a rat model of fetal alcohol exposure: dose-response and timing effects.
Alcohol 43:465–474.
Ivanov A, Pellegrino C, Rama S, Dumalska I, Salyha Y, Ben-Ari Y, Medina
I (2006) Opposing role of synaptic and extrasynaptic NMDA receptors in
regulation of the extracellular signal-regulated kinases (ERK) activity in
cultured rat hippocampal neurons. J Physiol 572:789–798.
Johnson TB, Goodlett CR (2002) Selective and enduring deficits in spatial
learning after limited neonatal binge alcohol exposure in male rats. Alcohol Clin Exp Res 26:83–93.
Kervern M, Silvestre de Ferron B, Alaux-Cantin S, Fedorenko O, Antol J,
Naassila M, Pierrefiche O (2015) Aberrant NMDA-dependent LTD after
perinatal ethanol exposure in young adult rat hippocampus. Hippocampus
25:912–923.
Klintsova AY, Helfer JL, Calizo LH, Dong WK, Goodlett CR, Greenough
WT (2007) Persistent impairment of hippocampal neurogenesis in young
adult rats following early postnatal alcohol exposure. Alcohol Clin Exp
Res 31:2073–2082.
Krapivinsky G, Krapivinsky L, Manasian Y, Ivanov A, Tyzio R, Pellegrino
C, Ben-Ari Y, Clapham DE, Medina I (2003) The NMDA receptor is coupled to the ERK pathway by a direct interaction between NR2B and
RasGRF1. Neuron 40:775–784.
Kuge T, Asayama T, Kakuta S, Murakami K, Ishikawa Y, Kuroda M, Imai
T, Seki K, Omoto M, Kishi K (1993) Effect of ethanol on the development
and maturation of synapses in the rat hippocampus: a quantitative electron-microscopic study. Environ Res 62:99–105.
Li C, Brake WG, Romeo RD, Dunlop JC, Gordon M, Buzescu R, Magarinos AM, Allen PB, Greengard P, Luine V, McEwen BS. (2004) Estrogen
GOODFELLOW ET AL.
alters hippocampal dendritic spine shape and enhances synaptic protein
immunoreactivity and spatial memory in female mice. Proc Natl Acad Sci
USA 101:2185–2190.
Lisman J, Raghavachari S (2015) Biochemical principles underlying the
stable maintenance of LTP by the CaMKII/NMDAR complex. Brain Res
1621:51–61.
Liu DD, Yang Q, Li ST (2013) Activation of extrasynaptic NMDA receptors
induces LTD in rat hippocampal CA1 neurons. Brain Res Bull 93:10–16.
Lu W, Man H, Ju W, Trimble WS, MacDonald JF, Wang YT (2001) Activation of synaptic NMDA receptors induces membrane insertion of new
AMPA receptors and LTP in cultured hippocampal neurons. Neuron
29:243–254.
Mameli M, Zamudio PA, Carta M, Valenzuela CF (2005) Developmentally
regulated actions of alcohol on hippocampal glutamatergic transmission.
J Neurosci 25:8027–8036.
Mattson SN, Crocker N, Nguyen TT (2011) Fetal alcohol spectrum disorders: neuropsychological and behavioral features. Neuropsychol Rev
21:81–101.
Monyer H, Burnashev N, Laurie DJ, Sakmann B, Seeburg PH (1994) Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 12:529–540.
Nixon K, Hughes PD, Amsel A, Leslie SW (2002) NMDA receptor subunit
expression following early postnatal exposure to ethanol. Dev Brain Res
139:295–299.
Nixon K, Hughes PD, Amsel A, Leslie SW (2004) NMDA receptor subunit
expression after combined prenatal and postnatal exposure to ethanol.
Alcohol Clin Exp Res 28:105–112.
Puglia MP, Valenzuela CF (2010) Repeated third trimester-equivalent ethanol exposure inhibits long-term potentiation in the hippocampal CA1
region of neonatal rats. Alcohol 44:283–290.
Quinn JJ, Loya F, Ma QD, Fanselow MS (2005) Dorsal hippocampus
NMDA receptors differentially mediate trace and contextual fear conditioning. Hippocampus 15:665–674.
Raybuck JD, Lattal KM (2014) Bridging the interval: theory and neurobiology of trace conditioning. Behav Processes 101:103–111.
Samudio-Ruiz SL, Allan AM, Sheema S, Caldwell KK (2010) Hippocampal
N-methyl-D-aspartate receptor subunit expression profiles in a mouse
model of prenatal alcohol exposure. Alcohol Clin Exp Res 34:342–353.
Schreiber WB, Hunt PS (2013) Deficits in trace fear conditioning induced by
neonatal alcohol persist into adulthood in female rats. Dev Psychobiol
55:352–361.
Shors TJ (1998) Stress and sex effects on associative learning: for better or
for worse. Neuroscientist 4:353–365.
Shouval HZ, Bear MF, Cooper LN (2002) A unified model of NMDA receptor-dependent bidirectional synaptic plasticity. Proc Natl Acad Sci USA
99:10831–10836.
Streissguth AP (2007) Offspring effects of prenatal alcohol exposure from
birth to 25 years: the Seattle Prospective Longitudinal Study. J Clin
Psychol Med Settings 14:81–101.
Tran TD, Kelly SJ (2003) Critical periods for ethanol-induced cell loss in the
hippocampal formation. Neurotoxicol Teratol 25:519–528.
Woolley CS, McEwen BS (1992) Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J Neurosci
12:2549–2554.
Woolley CS, Schwartzkroin PA (1998) Hormonal effects on the brain.
Epilepsia 39:2–8.
Yashiro K, Philpot B (2008) Regulation of NMDA receptor subunit expression and its implications for LTD, LTP, and metaplasticity. Neuropharmacology 55:1081–1094.
Zink M, Ferbert T, Frank ST, Seufert P, Gebicke-Haerter PJ, Spanagel R
(2011) Perinatal exposure to alcohol disturbs spatial learning and glutamate transmission-related gene expression in the adult hippocampus. Eur
J Neurosci 34:457–468.